Class KernelSVMOptions

java.lang.Object

org.tribuo.classification.sgd.kernel.KernelSVMOptions

All Implemented Interfaces:

com.oracle.labs.mlrg.olcut.config.Options, ClassificationOptions<KernelSVMTrainer>

public class KernelSVMOptions
extends Object
implements ClassificationOptions<KernelSVMTrainer>

Options for using the KernelSVMTrainer.

See:

 Shalev-Shwartz S, Singer Y, Srebro N, Cotter A
 "Pegasos: Primal Estimated Sub-Gradient Solver for SVM"
 Mathematical Programming, 2011.
 

Nested Class Summary

Nested Classes

Modifier and Type	Class	Description
static enum	KernelSVMOptions.KernelEnum	The kernel types.

Field Summary

Fields

Modifier and Type	Field	Description
double	kernelDegree	Degree in polynomial kernel function.
int	kernelEpochs	Number of SGD epochs.
double	kernelGamma	Gamma value in kernel function.
double	kernelIntercept	Intercept in kernel function.
KernelSVMOptions.KernelEnum	kernelKernel	Kernel function.
double	kernelLambda	Lambda value in gradient optimisation.
int	kernelLoggingInterval	Log the objective after n examples.
long	kernelSeed	Sets the random seed for the Kernel SVM.

Fields inherited from interface com.oracle.labs.mlrg.olcut.config.Options

header

Constructor Summary

Constructors

Constructor	Description
KernelSVMOptions()

Method Summary

Modifier and Type	Method	Description
KernelSVMTrainer	getTrainer()	Constructs the trainer based on the provided arguments.

Methods inherited from class java.lang.Object

clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Methods inherited from interface com.oracle.labs.mlrg.olcut.config.Options

getOptionsDescription

Field Details

kernelIntercept

@Option(longName="kernel-intercept",
        usage="Intercept in kernel function. Defaults to 1.0.")
public double kernelIntercept

Intercept in kernel function. Defaults to 1.0.

kernelDegree

@Option(longName="kernel-degree",
        usage="Degree in polynomial kernel function. Defaults to 1.0.")
public double kernelDegree

Degree in polynomial kernel function. Defaults to 1.0.

kernelGamma

@Option(longName="kernel-gamma",
        usage="Gamma value in kernel function. Defaults to 1.0.")
public double kernelGamma

Gamma value in kernel function. Defaults to 1.0.

kernelEpochs

@Option(longName="kernel-epochs",
        usage="Number of SGD epochs. Defaults to 5.")
public int kernelEpochs

Number of SGD epochs. Defaults to 5.

kernelKernel

@Option(longName="kernel-kernel",
        usage="Kernel function. Defaults to LINEAR.")
public KernelSVMOptions.KernelEnum kernelKernel

Kernel function. Defaults to LINEAR.

kernelLambda

@Option(longName="kernel-lambda",
        usage="Lambda value in gradient optimisation. Defaults to 0.01.")
public double kernelLambda

Lambda value in gradient optimisation. Defaults to 0.01.

kernelLoggingInterval

@Option(longName="kernel-logging-interval",
        usage="Log the objective after <int> examples. Defaults to 100.")
public int kernelLoggingInterval

Log the objective after n examples. Defaults to 100.

kernelSeed

@Option(longName="kernel-seed",
        usage="Sets the random seed for the Kernel SVM.")
public long kernelSeed

Sets the random seed for the Kernel SVM.

Constructor Details

KernelSVMOptions

public KernelSVMOptions()

Method Details

getTrainer

public KernelSVMTrainer getTrainer()

Description copied from interface: ClassificationOptions

Constructs the trainer based on the provided arguments.

Specified by:

getTrainer in interface ClassificationOptions<KernelSVMTrainer>

Returns:

The trainer.